Project S.A.L

Being a college student especially in Srishti, can get tough sometimes. We as students are expected to finish all our work on time, keep track of all the assignments given even though sometimes they’re posted on blackboard and we do not get any notifications regarding the same and then for when we do not end up finishing the work and decide to come late or are unable to make it to college it gets particularly hard to come up with the right “excuse” to tell the facilitator.
In order to help the students ease out their schedule and help them plan their work load more efficiently and researching within the context of N5 Campus in Srishti, we came up with the idea of S.A.L (Srishti’s Advanced Lockers), a conversational agent in the form of a screen embedded in each locker which helps the students organise their daily tasks and sets goals and targets for them!

After observing various spaces and objects around N5, all of us were sitting at home and brainstorming about what kind of issue we wanted to tackle and which aspect of N5 (which was our context for research), we wanted to take up. Mid-conversation, one of our friends who was doing her own work exclaims,

and that, just that sparked the idea of SAL. On questioning her as to why she wanted such a thing, she told us that she gets most of her assignments on Blackboard and sometimes she forgets to check blackboard and then is unable to do the required readings because it was in the contents section and no notification was provided later on and for her, it would be amazing to have someone set reminds for her on her phone without her having to check the date of every assignment and when it’s due. Plus, sometimes she needs a motivator to work. If somebody sets a task for her, she’ll have more motivation to do it than when she has to do it herself because that can result in a tonne of procrastination.
This got all of us thinking, if one student feels this one, many other might too. Within our group itself, majority of us could relate with her and hence, we decided to pursue the idea of redefining the lockers in Srishti to more than just abandoned storage spaces.

Lockers are something every student associates with a school or a college.
Any movie or a sitcom that we watch shows lockers in every school. “Locker Room Talk” for that matter, is an actual concept.
So, for this task we looked more into the state of the lockers in N5 campus and observed how they were being used. Most of the students walk past the lockers and the lockers for them have become more of an installation piece that nobody can touch rather than their own personal space. Some of the students use it as a mere storage space to dump their extra papers and the state of those lockers seemed really messy and abandoned. The emotional connect of the student and the locker lacked majorly and we wanted to bring that personalised connection back.

After observing the lockers we came to a few conclusions so as to why students around the campus might not use them as much -

SAL is a personal and fully customisable conversational agent embedded in the lockers of Srishti. It’s a screen in your locker.
It is a text based as well as sound based voice service that works at the touch of just a button! It gets activated by entering a password at the start of the day and then just by pressing the button at the bottom of the screen to activate the voice service throughout the day.
It is placed on the inner side of the door of the locker because that way the user is not disrupting the storage area and it also makes it easier for the user to use and type when its a text based service and speak after every button press when its a voice based service. It’s easy accessibility.
Since it’s fully customisable, you can change it’s name, alter any settings to your liking. You can make it a fully text based service as well and if there’s too much commotion and noise, it sends you a notification and converts to text based service itself.
A student’s identity information is provided to SAL when the lockers are assigned and blackboard access is given. It tracks the student’s daily affairs, syncs with their phone as well as blackboard and also works on the school wifi! SAL can help keep track of all the workload and classes that the student has. One can provide it their phone access so that SAL can help you set goals and targets by setting reminders for your assignments and help manage your work load. It’s also linked with the internet and the Srishti Main Server, it also has the ability to give you information about facilitators and send emails to them on your behalf.
It’s a machine learning device and adapts itself to the user. It knows what the user needs and it guides them through it or even does it for them making their life much more easier. SAL is a potential idea that makes the school storage lockers more than just storage spaces. It makes it a students own personal organiser and on top of that a pal.

In all honesty, there might not be a need for a conversational bot inside a locker but more of a want. Students have been managing to work and access Blackboard just fine but if there was something that could perform the same tasks for them quicker and more efficiently and also help them work better, why wouldn’t they want to try it?

Let’s try to tackle this with an example of a complex use case —
At the end of a college day a student goes back to dump all of his/her art supplies in the locker for the next studio session. After which while walking back home from college, throughout the way he/she is checking the work they have for that day by first, opening blackboard, entering the sign-in details, manoeuvring their way through the website, zooming in (if their accessing it on their phones), looking at the assignments and then settings goals (if they do) in their head, or in their phones or even could be by penning them down. All of this could be made so much more faster and efficient with the help of SAL. It would end at the first step itself, when a student does to put his/her art supplies in their lockers, they just activate the agent by asking about their assignments. Since, the agent is synced to the Srishti Main Server and also their blackboard accounts, it gives them a list of their assignments and also asks the student if he/she wants target setters for those assignments. If the student’s phone is synced with SAL then SAL automatically sets assignment targets for him/her on his/her calendar. The student can only change it and customise the targets to his/her liking. SAL acts as a motivator and makes the whole process much less of a task for the student.

We wanted to motivate the user to use the locker inserted of all of this conversation happening on their phone because on observation, Srishti lacks in providing students the personal connect and security with their own lockers and we wanted to bring that feeling back because a locker is a part of a school or college and a college without lockers feels incomplete but that’s where Srishti’s lockers were headed and we wanted to find a way to attract people and get more people to use their lockers.

We created a prototype video for our first assessment. It shows SAL being used . You can watch it here :

We mapped out how the conversation would flow and created a flow chart of the same.

We then moved onto creating a coded prototype of SAL using python3 taking some specific use cases and working on those exhaustively.

The uses cases chosen —

During various feedback sessions we were told that we were looking at a very linear and rigid way of conversation and that we should understand the complexities of conversation, so for our final coded prototype, we tried to add responses to follow up questions from the information that we provide to the user.

We were also told to optimise our code and try to lessen the use of nested if-else and try and incorporate more functions into the code and trying to solve this was a major issue for us, because of the way we had looped the code. If we would have added functions we would have had to change a lot of our code because of the way our looping constructs were working and so we decided against using functions, but to lessen the confusion we tried to comment our code fairly well.

After understanding the complexities of conversation, we tried to map out the conversation flow again, tying to lessen the binary and linear approach.

We have also created a demonstration of SAL showing the interaction of our coded prototype and a user. You can watch the video here :

Here is the live reaction of the user interacting with the coded prototype for the first time :

In the future we are planning to make SAL more conversational in its approach. We’re hoping to make it answer questions about happenings in and around N5.
Maybe, it could answer questions about upcoming events, help you navigate through N5. We are also aiming for our coded prototype to actually be able to set target goals for you an sync in with your phone, which is a functionality it doesn’t have yet.
We’re also hoping to add a lot more data into our coded prototype since it’s very limited for now and works only for the course “Human Centered Design” in Srishti.

After the interim exhibition and going through all the constructive feedback received, we listed down the problems and issues with our conversational agent —

Related posts:

Ihr erinnert euch bestimmt noch an die Studie, die ich euch im letzten Blogbeitrag vorgestellt habe. Es ging darum, welchen Einfluss die Musik auf das emotionale Befinden von Profimusikern hat…

Em algum momento da sua vida você já deve ter se imaginado ganhando dinheiro com tranquilidade, dormindo mais, tendo mais tempo pra você ou para sua família, não é mesmo? E se eu te dissesse que…

The first step in understanding your anger is recognizing that it’s a powerful emotion. Anger has the power to motivate you and help you take action. When people dismiss your anger, it’s often…